I've been out of IT for many years now, but one question I always have about these ransom scenarios is this: wouldn't advanced journaliing filesystems make recovery from an attack much easier, particularly filesystems where you can mount a shapshot? You could just start serving a past snapshot then make any updated files available as you clear them.

Back in the day I had customers who had incompetent DBAs bork their databases with bad SQL DML and DDL. Where the customer was using Oracle it was pretty easy to walk that stuff back because under the covers Oracle has been making heavy use of COW in their database storage. This allowed me to selectively walk back certain sets of problematic transactions. I could roll back just the transactions made by a certain user on a certain day that involved particular operations or database objects. You didn't have to figure out how to undo the individual effects of the bad transactions, you just waved your magic wand and it was as if those transactions never happened.

There must be some reason people aren't using file systems with COW and efficient snapshotting for general file service, because of on the face of it this seems like an obvious solution to the problem.

It's both easing of the accellerator and using the brakes to scrub off speed, with precise performance altered by the drive mode.

In this case the reasoning is somewhat circular. *If* there are many simulated worlds just like ours and there is only one real world, then it's more probable that our world is simulated than it is real. That's necessarily true, because it's a tautology. The truth of the statement as a whole tells you nothing about the world we actually live in.

As usual, tech bro hype has taken some impressive (to laymen) demos and spun them into a scenario that is far beyond was is demonstrably possible. Sure we can have the comptuer draw pretty pictures, but we actually can't model the world we live in very well. No computer model can tell you the price of Apple stock at the close of business tommorow or the temperature at 2PM in the afternoon a year from now. You can't model a fusion reactor sell enough to get to the point of building a working power station, you have to build many physical experiments to validate your model results. As the statistician George Box famously noted: all models are false; some models are useful.

As for faith, it has its place in science. You do an experiment because you feel confident it's going to tell you something; you usually have a pretty good idea of what you want to happen. That feeling of confidence is important in directing your efforts, but it carries no weight in arguments about results. Faith is only a "sin" (Greek *hamartia* -- to miss the mark) when you demand others share it.

For most people some of the time, and for just about everyone some of the time, modern automatic transmissions will perform better than they would with an ICE vehicle. But no matter how good any automatic transmission is, the one thing it will never be able to do is read your mind about what you *intend* to do next. So there will always be situations with an ICE vehicle where you'd rather have a manual or semi-automatic than an automatic.

That doesn't apply to electric motors, which produce nearly peak torque at 0 RPM and then over a wide range of RPMs; so you never have to match the motor's RPMs to what you want to do next. There are corner cases, like towing an extremely heavy load or traveling at extremely high speeds outside the motor's very wide power band, where you'd want to have different gear ratios. There are various ways for engineers to address these cases, but if they chose to give a vehicle a shiftable transmission, there's no reason that a computer couldn't do the shifting; there's no need for it to "read your mind".

As for on snow, regenerative braking can feel a lot like engine braking depending on your driving settings. In a vehicle's maximum efficiency mode the motor will very noticeably begin to absorb energy from the wheels when you let up on the "gas".

A chainsaw is a great tool. Some fool chopping his leg off with one doesn't change that.

Yeah, the summary was missing a conditional clause regarding the intent of the action, to prevent repair with an otherwise functional part.

If serial number anti-removal techniques are working to prevent repair now and in the future, they will also prevent theft now and in the same time frame of the future.

The sophistication of high-tier determined thieves is about the same as professional independent repair shops. The main issue a non-thief independent-repair-shop may need to overcome is whether or not it is illegal to circumvent the protection, which things like the DMCA aren't supposed to touch even now.

What the parent poster was proposing is that registration of the same serial number as already sold may be blocked, but a non-serialized or other-serialized party would be allowed. Currently the restrictions are designed to require a match in a central registry. The central registry if it exists can simply be made to allow new things not linked to a registered owner by the manufacturer already.

And of course, for vehicles and guns we use vin and serial number etching which is manipulable too, but it is still effective at least somewhat in tracking crime. I suspect things would be similar for such electronically etched systems.

And these electronically etched systems generally work by having a writable memory that can have its write system "burned" by triggering a "fuse" that permanently makes it impossible to rewrite it without destruction or wholesale replacement of the chip itself. The goal here is to protect the chip, which is what is generally proprietary here. This also works for firmware protection too, to some extent, unless you have some especially powerful equipment.

Oregon citizen and friend of Charlie (who was quoted in the summary) here.

The bill specifically bans parts pairing when made effective to prevent an otherwise functional part from being used for replacement.

You can still use part serial numbering to prevent theft.

That's just getting old. When you're middle-aged you're too busy for things that used to give you joy, but you spend years lugging them around with the vague idea you'll get back to them someday. That includes people too; in middle age you don't priortize the people who make you happy, you prioritize the ones you need to get through your busy days. Then if you're old enough, you'll find yourself with an addressbook somewhere with dozens of numbers of people you'd like to call but with numbers that probably haven't worked in years.

Losing touch with something or someone doesn't necessarily mean that thing or person wasn't worthwhile. Sure, if you're 70 you should probably get rid of that ice-axe or scuba gear you haven't used in forty years. But there's probably things you threw out that you wish you still had.

I'm not saying it *can't* be someone who is enthusiastic about aviation, but there's the danger of what psychologists call "motivated thinking" -- you or I would call it "wishful thinking" or "denial". If someone really loves the product, you need him to be able to believe something he wish was not true.

That's actually a rare quality. If a close call by a referee goes against your team, I'd say at least 90% of people would automatically believe it was a bad call and could not be convinced otherwise.

No, you don't want someone who *romanticizes* aviation. You want a hard-nosed realist who can think critically and has got his priorities straight.

In a crisis of trust, what you need is a leader with *character*. You need someone who understands the responsibility of building a product that people trust, but which can kill them. You need someone who can speak with discretion while at the same time being scrupulously honest about things people have a right to know.

Above all, you need a realist who is going focus on things that make an actual difference rather than just managing perceptions and evading blame. Saying the right thing has never been Boeing's management's problem, they always said exactly what needed to be said. They just never did what needed to be done.

I think the need for a leader with outstanding character is why people would like to see an engineer in charge. People trust engineers, otherwise they'd never get in a car or a plane. But we're not all ideal engineers, are we? Look at the CEO of OceanGate; he was an aeronautical engineer, but you sure as hell wouldn't want someone like him in charge of Boeing. A lack of enthusiasm for his vehicle wasn't what caused it to implode; it was a lack of sober and critical thinking.

That's really the only thing I miss about vinyl -- Thos great big album covers. Especially the two disc sets, which had covers that opened like books, or even were books.

Anyone who doesn't remember this can get an idea of what it was like here.

Large corporations aren't *national*, and calling them "multi-national" doesn't really capture their nature which isn't isn't subordinate to any nation, much less multiple ones. They stand outside the framework of nations, acting as power and interest centers entirely in their own right.

The only way to get a large corporation to respect the interests of a nation-state is for that nation-state to compel that by force -- through laws, regulations, and in authoritarian states extrajudicial punishments. The only reason for Americans to be *particularly* concerned about TikTok is that it particularly convenient for the Chinese government to enforce its will on them. In fact if you look at ByteDance's response -- offering to move information centers outside of China, you will see an example of a corporation not identifiying with its host country's interests and trying to circumvent state control.

Now vis-a-vis China, the Xi regime is attempting to exert more control over businesses operating in their countries *which is deterimental to the interests of multi-national corporations operating there*. And in any case foreign social media sites aren't allowed to operate in the country. So we don't have to worry about, say, Meta getting too cozy with Xi; The Xi regime is Meta's *enemy*. But you can bet Meta would sell you to the Chinese government if the regime were more friendly, and even to respond positively to friendly overtures.

Chinese people aren't identical, ideological robots, whether you believe that to be the product of race, culture, or authoritarian training. Sure, culture and indoctrination at least may predispose *more* people there to take a Maoist or nationalist viewpoint, but I guarantee you that among 1.4 billion people there are plenty dissidents and freethinkers and people who question the party line. Xi is really reimposing elements of Maoism after a decades long turn away from that, which means he has to purge *even the CCP party ranks* of diversity.

If you couldn't expect any diversity of viewpoint in China, the the CCP would not be so keen on cracking down on dissent. They wouldn't be banging the "Xi Xinping Thought" drum so loudly if there weren't other viewpoints out there that needed to be drowned out.

The idea that Mao "won" because he had some kind of mystically infallible and exclusive insight into the character of Chinese people is a "just so" story -- reasoning that becaue he won he *must* have had some special understanding of China. Not to unduly minimize his perspicacity as a politician, he won beause this opposition -- the KMT -- was really, really bad. Mao was not harnessing some kind of reactionary Chinese impulse to resist an attempt to impose western economic and political liberalism; communism was a modern alternative to a reacdtionary and nationalistic regime whose corruption and incompetence had led the nation to high unemployment rates and hyperinflation.

As the Italians say, "traduttore traditore." I'd translate what that means into Engish but it loses something.